Question

Your team maintains separate Azure Machine Learning workspaces for the development, test, and production environments, each in a different resource group. A data science team has trained and registered a model in the development workspace. You must make this exact model, along with its associated environment and components, available for deployment in the test and production workspaces without retraining or manually copying files between workspaces. What should you do?

Accepted Answer

Create an Azure Machine Learning registry and promote the model, environment, and components to the registry, then reference them from the test and production workspaces.

Answer

Register the model as a data asset in each workspace and point the data asset URI to the development workspace's blob storage.

Answer

Configure the development workspace datastore as a shared datastore in the test and production workspaces so they can read the model files directly.

Answer

Export the model using the Azure CLI and re-register it manually in each target workspace with the same name and version.